21 views

Uploaded by Amit Shukla

Minimum variance feedback

- The Matrix Cookbook
- EXPLICIT POLAR DECOMPOSITION OF COMPANION MATRICES
- Matrices and Determinants Worksheet 1
- a Practical Approach to Linear Algebra
- Honors Advanced Calculus and Linear Algebra pp
- The Matrix Cookbook
- Generalized state estimation
- Discrete Mathematics Questions
- Matrix Primer
- Mathematics I
- Op Tim Ization Problem
- Basic Statistics
- DirectMethod-TrussAnalysis
- A Matrix Pseudo-Inversion Lemma
- Simulation and Control
- epdf.pub_sampling-and-statistical-methods-for-behavioral-ec.pdf
- lu
- matrix change basisbasis
- W6 MATLAB Fundamentals
- State Estimation

You are on page 1of 4

www.elsevier.com/locate/automatica

Technical communique

discrete-time systems with direct feedthrough

Steven Gillijns , Bart De Moor

SCD-SISTA, ESAT, K.U.Leuven, Kasteelpark Arenberg 10, 3001 Leuven, Belgium

Received 17 May 2006; received in revised form 29 September 2006; accepted 21 November 2006

Available online 13 March 2007

Abstract

This paper extends previous work on joint input and state estimation to systems with direct feedthrough of the unknown input to the

output. Using linear minimum-variance unbiased estimation, a recursive lter is derived where the estimation of the state and the input are

interconnected. The derivation is based on the assumption that no prior knowledge about the dynamical evolution of the unknown input is

available. The resulting lter has the structure of the Kalman lter, except that the true value of the input is replaced by an optimal estimate.

2007 Elsevier Ltd. All rights reserved.

Keywords: Kalman ltering; Recursive state estimation; Unknown input estimation; Minimum-variance estimation

1. Introduction

Systematic measurement errors and model uncertainties such

as unknown disturbances or unmodeled dynamics can be represented as unknown inputs. The problem of optimal ltering

in the presence of unknown inputs has therefore received a lot

of attention.

Friedland (1969) and Park, Kim, Kwon, and Kwon (2000)

solved the unknown input ltering problem by augmenting

the state vector with an unknown input vector. However, this

method is limited to the case where a model for the dynamical

evolution of the unknown input is available.

A rigorous and straightforward state estimation method in the

presence of unknown inputs is developed by Hou and Mller

(1994) and Hou and Patton (1998). The approach consists in

rst building an equivalent system which is decoupled from

the unknown inputs, and then designing a minimum-variance

unbiased (MVU) estimator for this equivalent system.

This paper was not presented at any IFAC meeting. This paper was

recommended for publication in revised form by Associate Editor Karl Henrik

Johansson under the direction of Editor Andr Tits.

Corresponding author. Tel.: +32 16 32 17 09; fax: +32 16 32 19 70.

E-mail addresses: steven.gillijns@esat.kuleuven.be (S. Gillijns),

bart.demoor@esat.kuleuven.be (B. De Moor).

0005-1098/$ - see front matter 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.automatica.2006.11.016

Another approach consists in parameterizing the lter equations and then calculating the optimal parameters by minimizing the trace of the state covariance matrix under an unbiasedness condition. An optimal lter of this type was rst developed by Kitanidis (1987). The derivation of Kitanidis (1987)

is limited to linear systems without direct feedthrough of the

unknown input to the output and yields no estimate of the

input. An extension to state estimation for systems with direct feedthrough was developed by Darouach, Zasadzinski, and

Boutayeb (2003). Extensions to joint input and state estimation

for systems without direct feedthrough are addressed by Hsieh

(2000) and Gillijns and De Moor (2007).

In this paper, we combine both extensions of Kitanidis (1987)

by addressing the problem of joint input and state estimation

for linear discrete-time systems with direct feedthrough of the

unknown input to the output. Using linear minimum-variance

unbiased estimation, we develop a recursive lter where the

estimation of the state and the input are interconnected. The

estimation of the input is based on the least-squares (LS) approach developed by Gillijns and De Moor (2007), while the

state estimation problem is solved using the method developed

by Kitanidis (1987).

This paper is outlined as follows. In Section 2, we formulate

the ltering problem and present the recursive three-step structure of the lter. Next, in Sections 35, we consider each of

the three steps separately and derive equations for the optimal

input and state estimators. Finally, in Section 6, we summarize

the lter equations.

935

5, where we calculate the optimal value of Lk .

3. Time update

2. Problem formulation

Consider the linear discrete-time system

xk+1 = Ak xk + Gk dk + wk ,

yk = Ck xk + Hk dk + vk ,

(1)

(2)

vector, and yk Rp is the measurement. The process noise

wk Rn and the measurement noise vk Rp are assumed

to be mutually uncorrelated, zero-mean, white random signals

with known covariance matrices, Qk = E[wk wkT ] 0 and Rk =

E[vk vkT ] > 0, respectively. Results are easily generalized to the

case where wk and vk are correlated by applying a preliminary

transformation to the system (Anderson & Moore, 1979). Also,

results are easily generalized to systems with both known and

unknown inputs. The matrices Ak , Gk , Ck and Hk are known

and it is assumed that rank Hk = m. Throughout the paper, we

assume that (Ak , Ck ) is observable and that x0 is independent of

vk and wk for all k. Also, we assume that an unbiased estimate

x0 of the initial state x0 is available with covariance matrix P0x .

The objective of this paper is to design an optimal recursive

lter which estimates both the system state xk and the input

dk based on the initial estimate x0 and the sequence of measurements {y0 , y1 , . . . , yk }. No prior knowledge about the dynamical evolution of dk is assumed to be available and no prior

assumption is made. The unknown input can be any type of

signal.

The optimal state estimation problem for a system with direct feedthrough of the unknown input dk to the output yk is

conceptually not very different from the case where Hk = 0.

A single lter and a single existence condition, valid for both

cases, can be found in Darouach et al. (2003) and Hou and

Mller (1994). In contrast, the optimal input estimation problem is conceptually very different in both cases. If Hk = 0,

the unknown input dk must be estimated with one step delay

because the rst measurement containing information on dk is

yk+1 (Gillijns & De Moor, 2007). On the other hand, if Hk = 0,

the rst measurement containing information on dk is yk . Consequently, the structure of the input estimator and the existence

conditions are totally different in both cases.

We consider a recursive three-step lter of the form

n

dk = Mk (yk Ck xk|k1 ),

xk|k = xk|k1 + Lk (yk Ck xk|k1 ),

(3)

(4)

(5)

be determined. The rst step, which we call the time update,

yields an estimate of xk given measurements up to time k 1.

This step is addressed in Section 3. The second step yields an

estimate of the unknown input. The calculation of the optimal

matrix Mk is addressed in Section 4. Finally, the third step, the

so-called measurement update, yields an estimate of xk given

denote the optimal unbiased estimates of xk1 and dk1 given

measurements up to time k 1, then the time update is given by

xk|k1 = Ak1 xk1|k1 + Gk1 dk1 .

The error in the estimate xk|k1 is given by

xk|k1 := xk xk|k1 ,

= Ak1 xk1|k1 + Gk1 dk1 + wk1 ,

with xk|k := xk xk|k and dk := dk dk . Consequently, the

covariance matrix of xk|k1 is given by

x

:= E[xk|k1 x Tk|k1 ],

Pk|k1

x

P

= [ Ak1 Gk1 ] k1|k1

dx

Pk1

xd

Pk1

d

Pk1

ATk1

+ Qk1 ,

GTk1

x

:= E[xk|k x Tk|k ], Pkd := E[dk d Tk ] and (Pkxd )T =

with Pk|k

dx

Pk := E[dk x Tk|k ]. Expressions for these covariance matrices

will be derived in the next sections.

4. Input estimation

In this section, we consider the estimation of the unknown

input. In Section 4.1, we determine the matrix Mk such that (4)

yields an unbiased estimate of dk . In Section 4.2, we extend to

MVU input estimation.

4.1. Unbiased input estimation

Dening the innovation yk := yk Ck xk|k1 , it follows

from (2) that

yk = Hk dk + ek ,

(6)

where ek is given by

ek = Ck xk|k1 + vk .

(7)

consequently from (6) that E[yk ] = Hk E[dk ]. This indicates that

an unbiased estimate of the unknown input dk can be obtained

from the innovation yk .

Theorem 1. Let xk|k1 be unbiased, then (3)(4) is an unbiased

estimator for all possible dk if and only if Mk satises Mk Hk =I .

Proof. The proof is similar to that of Theorem 1 in Gillijns

and De Moor (2007) and is omitted.

It follows from Theorem 1 that rank Hk = m is a necessary and sufcient condition for the existence of an unbiased

936

input estimator of the form (4). Note that this condition implies p m. The matrix Mk = (HkT Hk )1 HkT corresponding to

the LS solution of (6) satises the condition of Theorem 1.

The LS solution is thus unbiased. However, it follows from the

GaussMarkov theorem (Kailath, Sayed, & Hassibi, 2000) that

it is not necessarily minimum-variance because in general

x

R k := E[ek ekT ] = Ck Pk|k1

CkT + Rk = cI ,

Proof. We use the approach of Kitanidis (1987), where a similar optimization problem is solved using Lagrange multipliers.

The Lagrangian is given by

x

x

trace{Lk R k LTk 2Pk|k1

CkT LTk + Pk|k1

}

An MVU estimate of dk based on the innovation yk is obtained by weighted LS estimation with weighting matrix equal

to the inverse of R k .

Theorem 2. Let xk|k1 be unbiased and let R k and HkT R 1

k Hk

be nonsingular, then for Mk given by

1 T 1

Mk = (HkT R 1

k Hk ) Hk R k ,

optimal input estimate is given by

1

Pkd = (HkT R 1

k Hk ) .

and De Moor (2007) and is omitted.

We denote the optimal input estimate corresponding to Mk

by d k and derive an equation for d k := dk d k . It follows

from (4), (6) and the unbiasedness of the input estimator that

d k is given by

d k = (I Mk Hk )dk Mk ek = Mk ek .

(8)

This equation will be used in the next section, where we consider the measurement update.

5. Measurement update

Finally, we consider the update of xk|k1 with the measurement yk . We calculate the gain matrix Lk which yields the

MVU estimator of the form (5). Using (5) and (6), we nd that

xk|k = (I Lk Ck )xk|k1 Lk Hk dk Lk vk .

(9)

Lk satises

Lk Hk = 0.

(10)

x

Let Lk satisfy (10), then it follows from (9) that Pk|k

is given by

x

x

Pk|k

= (I Lk Ck )Pk|k1

(I Lk Ck )T + Lk Rk LTk .

(11)

gain matrix Lk which minimizes the trace of (11) under the

unbiasedness condition (10).

(12)

x

where Kk = Pk|k1

CkT R 1

k , minimizes the trace of (11) under

the unbiasedness condition (10).

2 trace{Lk Hk Tk },

(13)

the factor 2 is introduced for notational convenience. Setting

the derivative of (13) with respect to Lk equal to zero, yields

x

R k LTk Ck Pk|k1

Hk Tk = 0.

T

x

Lk

R k Hk

Ck Pk|k1

=

,

0

HkT

0

Tk

(14)

(15)

which has a unique solution if and only if the coefcient matrix is nonsingular. Let R k be nonsingular, then the coefcient

matrix is nonsingular if and only if HkT R 1

k Hk , the Schur com

plement of Rk , is nonsingular. Finally, premultiplying left- and

right-hand side of (15) by the inverse of the coefcient matrix,

yields (12).

We denote the state estimate corresponding to the gain matrix

Lk by x k|k . Substituting (12) in (5), yields the equivalent state

updates

x k|k = xk|k1 + Kk (I Hk Mk )(yk Ck xk|k1 ),

= xk|k1 + Kk (yk Ck xk|k1 Hk d k ),

from which we conclude that the optimal state estimator implicitly estimates the unknown input by weighted LS estimation.

Finally, we derive expressions for the covariance matrices

x := E[x x T ] and P xd := E[x d T ] where

Pk|k

k|k k|k

k|k k

k

x k|k := xk x k|k ,

= (I Lk Ck )xk|k1 Lk vk .

(16)

x ,

for Pk|k

x

x

Pk|k

= Pk|k1

Kk (R k Hk Pkd HkT )KkT .

x

Pkxd = Pk|k1

CkT MkT = Kk Hk Pkd .

In this section, we summarize the lter equations. We assume

that x0 , the estimate of the initial state, is unbiased and has

given by:

Initialization:

P0x = E[(x0 x0 )(x0 x0 )T ].

The recursive part of the lter consists of three steps: the estimation of the unknown input, the measurement update and the

time update. These three steps are given by

Estimation of unknown input:

x

R k = Ck Pk|k1

CkT + Rk ,

Mk = (HkT R 1 Hk )1 HkT R 1 ,

k

dk = Mk (yk Ck xk|k1 ),

P d = (HkT R 1 Hk )1 .

Measurement update:

x

CkT R 1

Kk = Pk|k1

k ,

Px = Px

Kk (R k Hk P d HkT )KkT ,

k|k

Pkxd

where the optimal weighting matrix is computed from the covariance matrices of the state estimator.

Acknowledgments

x0 = E[x0 ],

937

k|k1

= (Pkdx )T

Our research is supported by Research Council KULeuven: GOA AMBioRICS, several PhD/postdoc & fellow

Grants; Flemish Government: FWO: PhD/postdoc Grants,

projects, G.0407.02 (support vector machines), G.0197.02

(power islands), G.0141.03 (Identication and cryptography),

G.0491.03 (control for intensive care glycemia), G.0120.03

(QIT), G.0452.04 (new quantum algorithms), G.0499.04

(Statistics), G.0211.05 (Nonlinear), research communities (ICCoS, ANMMM, MLDM); IWT: Ph.D. Grants, GBOU (McKnow); Belgian Federal Science Policy Ofce: IUAP P5/22

(Dynamical Systems and Control: Computation, Identication and Modelling, 20022006); PODO-II (CP/40: TMS

and Sustainability); EU: FP5-Quprodis; ERNSI; Contract Research/agreements: ISMC/IPCOS, Data4s, TML, Elia, LMS,

Mastercard.

References

= Kk Hk Pkd .

Time update:

xk+1|k = Ak xk|k + Gk dk ,

x

P

x

Pk+1|k

= [ Ak Gk ] k|k

Pkdx

Pkxd

Pkd

ATk

+ Qk .

GTk

Note that the time and measurement update of the state estimate take the form of the Kalman lter, except that the true

value of the input is replaced by an optimal estimate. Also, note

that in case Hk = 0 and Gk = 0, the Kalman lter is obtained.

7. Conclusion

This paper has studied the problem of joint input and

state estimation for linear discrete-time systems with direct

feedthrough of the unknown input to the output. A recursive

lter was developed where the update of the state estimate has

the structure of the Kalman lter, except that the true value of

the input is replaced by an optimal estimate. This input estimate

is obtained from the innovation by weighted LS estimation,

Cliffs, NJ: Prentice-Hall.

Darouach, M., Zasadzinski, M., & Boutayeb, M. (2003). Extension

of minimum variance estimation for systems with unknown inputs.

Automatica, 39, 867876.

Friedland, B. (1969). Treatment of bias in recursive ltering. IEEE

Transactions on Automatic Control, 14, 359367.

Gillijns, S., & De Moor, B. (2007). Unbiased minimum-variance input

and state estimation for linear discrete-time systems. Automatica, 43(1),

111116.

Hou, M., & Mller, P. C. (1994). Disturbance decoupled observer design:

A unied viewpoint. IEEE Transactions on Automatic Control, 39(6),

13381341.

Hou, M., & Patton, R. J. (1998). Optimal ltering for systems with unknown

inputs. IEEE Transactions on Automatic Control, 43(3), 445449.

Hsieh, C. S. (2000). Robust two-stage Kalman lters for systems with

unknown inputs. IEEE Transactions on Automatic Control, 45(12),

23742378.

Kailath, T., Sayed, A. H., & Hassibi, B. (2000). Linear estimation. Upper

Saddle River, NJ: Prentice-Hall.

Kitanidis, P. K. (1987). Unbiased-minimum variance linear state estimation.

Automatica, 23(6), 775778.

Park, S. H., Kim, P. S., Kwon, O., & Kwon, W. H. (2000). Estimation and

detection of unknown inputs using optimal FIR lter. Automatica, 36,

14811488.

- The Matrix CookbookUploaded byDhillonv
- EXPLICIT POLAR DECOMPOSITION OF COMPANION MATRICESUploaded byYhuda Septiawan
- Matrices and Determinants Worksheet 1Uploaded byAravindh Shankar
- a Practical Approach to Linear AlgebraUploaded byDeepak Singh
- Honors Advanced Calculus and Linear Algebra ppUploaded bymarchelo_chelo
- The Matrix CookbookUploaded by艾弗
- Generalized state estimationUploaded byAnonymous 1880JHcG
- Discrete Mathematics QuestionsUploaded bySaravana Suresh Gomathi
- Matrix PrimerUploaded bymir emmett
- Mathematics IUploaded bynipuna22
- Op Tim Ization ProblemUploaded byEmin Bekir Altunbaş
- Basic StatisticsUploaded byAbhinav Khare
- DirectMethod-TrussAnalysisUploaded bymeghatv
- A Matrix Pseudo-Inversion LemmaUploaded byines_elleuch
- Simulation and ControlUploaded bysolid34
- epdf.pub_sampling-and-statistical-methods-for-behavioral-ec.pdfUploaded byPriyanka ghosh
- luUploaded bymikeljosep
- matrix change basisbasisUploaded byMilli Yeshiwas
- W6 MATLAB FundamentalsUploaded byDanny Manno
- State EstimationUploaded byPedro Nuñez Suarez
- Buss Maths Syllabs Funcs-QEqu-Maxima EctUploaded byRaghunath Rangaraju
- jemh108Uploaded byAnderson Alfred
- 4. Material I_EEE 403 ACT Part 2.pdfUploaded byNishant Kamal
- AET Math Camp 2018.pdfUploaded byLee RickHunter
- mm-29Uploaded bySomasekhar Chowdary Kakarala
- Computer ScienceUploaded byAnurag Garg
- 6326143Uploaded byAndika Surya
- Colgate 1993 RobustImpedanceShapingTelemanipulationUploaded byverdos
- Calc15.8Uploaded byJeoff Libo-on
- Lecture 1Uploaded byFarhana Aziz

- Power ElectronicsUploaded byYugendra Rao
- Artificial Intelligent System for Steam Boiler Diagnosis Based on Superheater MonitoringUploaded byAmit Shukla
- Power Electronics (Jameel Asghar)Uploaded byAhmed Rehan
- Chapter 1_1 - 80pUploaded byAmit Shukla
- Development of autonomous manoeuvres in a quadcopterUploaded bymarcsure
- State and Unknown Input Estimation for Linear Discrete-time SystemsUploaded byAmit Shukla
- Chapter 1_ 27Uploaded byAmit Shukla
- Peshwa Bajirao I and Maratha ExpansionUploaded byAmit Shukla
- MMEG221_Project5Uploaded byAmit Shukla
- Vaisnava CalendarUploaded byAmit Shukla
- Management Quadrocopters ArDrone 2Uploaded byamitshukla.iitk
- 21 Hindu Temples Outside IndiaUploaded byAmit Shukla
- MMEG221_Project2Uploaded byAmit Shukla
- Lab10Uploaded byAmit Shukla
- संन्धिUploaded byAmit Shukla
- sheet2Uploaded byAmit Shukla
- Advaitavada HindiUploaded byAmit Shukla
- Vaisnava Calendar LucknowUploaded byAmit Shukla
- State and Input Estimation for Descriptor Systems With Unknown InputsUploaded byAmit Shukla
- Unbiased Minimum-Variance Input and State Estimation for Linear Discrete-time SystemsUploaded byAmit Shukla
- MMEG221_Project3Uploaded byAmit Shukla
- Lect SplineUploaded byAmit Shukla
- Unbiased Minimum-Variance Linear State EstimationUploaded byAmit Shukla
- Matrices in Matlab Lecture8Uploaded byAmit Shukla
- MMEG221_Project4Uploaded byAmit Shukla
- Origins of Such Languages as SanskritUploaded byAmit Shukla
- Leak Detection in Water PipeUploaded byAmit Shukla
- Seekur JRUploaded byAmit Shukla
- Seekur Large SizeUploaded byAmit Shukla

- Translate Lampiran IIUploaded byganedn
- Omaree Autopsy ReportUploaded byKOB4
- The Descent of the Dove - Charles Williams (pagination corrected)Uploaded byWF1900
- 3 Distance Measuring (Hurdle Detection System)Uploaded byShweta Ramteke
- 8BL45GB - ManualUploaded byJorge Chapi
- Pengkajian Keperawatan Pada AnakUploaded byWafa Aun'k
- Gilles Deluze the Movement ImageUploaded byPetar Dekic
- Technology Resources GuideUploaded byvlr001
- EIL Material Selection ChartUploaded byVijaya Sekhar
- CobasUploaded byأبو عبد الله المصري
- S1-2016-313016-bibliographyUploaded byBintang Alfian Nur Rachman
- englez - roman.txtUploaded bygerte_yuew
- Physics Dotpoint Hscphysics Q&AUploaded bymeli
- Correlation and Path Analysis in Yield and Yield Components in Spring Bread Wheat (Triticum Aestivum L.) Genotypes Under Irrigated Condition in Southern IndiaUploaded byIsmael Neu
- GST100 Installation and operation ManualUploaded byetchegarayf
- Eigen Value Extraction MethodUploaded byjy4661
- DNYANESHWARIUploaded byAmol Modak
- The Quadcopter _ How to Compute the Pitch, Roll and Yaw – TheboredengineersUploaded byanton
- Ce2257 Lab ManualUploaded byNiranjan
- Brita Horn Steamboat PilotUploaded byBrita Horn
- Eugene Butkov PDFUploaded byBrian
- graphing trig functionsUploaded byapi-285179261
- Primer TimelineUploaded byGentle1337
- Nag Hammadi Library - MelchizedekUploaded byDavid Jose Lopez Wong
- Multilevel Voltage-Source-Converter Topologies for Industrial Medium-Voltage DrivesUploaded byVamsi Sakhamuri
- Matlab_6 (1).pdfUploaded byHana Hamid
- TL MR3420 V3 UGUploaded bydol_tpc
- Form 2 Science Mock Test (Sem 1 Term 1)Uploaded bySpike Chingyen
- Transport-Documents-Used-In-International-Trade.pdfUploaded bynguyentamvnuk
- Huaxin Antenna BrochureUploaded byFmwingenieriasa Maroja